Abstract 4370153: Late gadolinium enhanced cardiac magnetic resonance imaging scar is associated with a delayed effectiveness of radiofrequency ablation of premature ventricular contractions.

Abstract

Background: Radiofrequency ablation (RFA) is an effective treatment of premature ventricular contractions (PVC). Patients with an initially unsuccessful ablation may demonstrate later suppression of PVCs; the incidence and mechanism of late supression is not understood. The presence of cardiac scar modulates tissue response to ablation and may be associated with this phenomenon. Objective: To compare patients with acutely successful ablation procedures to patients with delayed effective ablations, and to assess the impact of late-gadolinium enhanced cardiac magnetic resonance imaging (LGE-CMR) on this occurrence. Methods: Patients undergoing PVC ablation with LGE-CMR were included. All patients underwent ablation with the endpoint of elimination of the PVC or non-capture at high-output pacing (20mv and 10ms) at the site of origin and/or anatomically adjacent sites. All patients were admitted for observation and had assessment of PVC burden prior to discharge and at 3 month follow up. Successful procedures (80% PVC burden reduction) was seen at the completion of ablation (acute success), within 24 hours (early delayed success) or at 3 months follow up (late delayed success). Results: Among 343 patients, 278(81%) had successful procedures (age 53±15, male n=141(51%), ejection fraction 50±12, PVC burden 20±13, non-ischemic cardiomyopathy(NICM) n=30(11%), ischemic cardiomyopathy (n=28(10%), LGE-CMR scar n=48(20%)), including patients with acute success (n=254), early-delayed success (n=14), or late-delayed success (n=10). Patients with delayed success (n=24) had greater rates of LGE-CMR scar (55% vs 16%), lower EFs (44±12% vs 51±12%), NICM (42% vs 8%), and intramural PVCs (67% vs 17%) (P<0.001 for all), with no other demographic differences. Four/24 patients with a delayed effect were on anti-arrhythmic medications at follow up, including 2 who had previously failed amidoarone. LGE-CMR was associated with a delayed effective procedure independent of the age, sex, EF or NICM (OR 3.9 95%CI [1.2-12.6], P=0.02). Conclusion: A delay in the effectiveness of RFA for PVCs was seen in 9% of patients, and manifested as early as 24 hours post ablation in 58% of those patients. The presence of LGE-CMR scar was independently associated with a delay in RFA effectiveness. A strategy of ablation to non-capture at high-output pacing and delay of additional PVC treatment until clinic follow up should be considered in patients with an initially unsuccessful ablation procedure.